11119788

Serialization Floors and Deadline Driven Control for Performance Optimization of Asymmetric Multiprocessor Systems

PublishedSeptember 14, 2021
Assigneenot available in USPTO data we have
Technical Abstract

Patent Claims
12 claims

Legal claims defining the scope of protection, as filed with the USPTO.

1

1. A method of controlling performance of one or more processors or coprocessors of an asymmetric multiprocessor system, the method comprising: executing a thread group on a processor and a coprocessor of the asymmetric multiprocessor system, wherein the thread group has a first control effort parameter corresponding to the processor and a second control effort parameter corresponding to the coprocessor; and at least one of: performing a hysteretic adjustment of the first control effort parameter to slow a transition of the processor to a low power state while a workload associated with the thread group is executing on the coprocessor; or performing a hysteretic adjustment of the second control effort parameter to slow a transition of the coprocessor to a low power state while the workload associated with the thread group is executing on the processor.

2

2. The method of claim 1 wherein the hysteretic adjustment of at least one of the first control effort parameter and the second control effort parameter includes delaying a time between the workload being submitted to the coprocessor and decreasing the first control effort parameter.

3

3. The method of claim 1 wherein the hysteretic adjustment includes decreasing a rate at which the first control effort parameter decreases.

4

4. The method of claim 1 wherein the processor is selected from the group consisting of a central processing unit, a graphics processing unit, a general purpose graphics processing unit, a neural engine, an image signal processor, and a scaling and rotating engine.

5

5. The method of claim 1 wherein the coprocessor is selected from the group consisting of a central processing unit, a graphics processing unit, a general purpose graphics processing unit, a neural engine, an image signal processor, and a scaling and rotating engine.

6

6. The method of claim 1 wherein the first and second control effort parameters respectively affect at least one of an allocated subset of cores or execution units and a dynamic voltage and frequency state of the processor or the coprocessor.

7

7. An asymmetric multiprocessor system comprising: a processor complex comprising one or more processors; one or more coprocessors; a closed loop performance controller configured to control performance of the one or more processors and the one or more coprocessors; and an operating system executing on the processor complex, the operating system comprising an input/output service interactive with the closed loop performance controller and one or more drivers corresponding to the one or more coprocessors; wherein the closed loop performance controller is configured to cooperate with the operating system, the processor complex, and the one or more coprocessors to: execute a thread group on a processor and a coprocessor of the asymmetric multiprocessor system, wherein the thread group has a first control effort parameter corresponding to the processor and a second control effort parameter corresponding to the coprocessor; and at least one of: perform a hysteretic adjustment of the first control effort parameter to slow a transition of the processor to a low power state while a workload associated with the thread group is executing on the coprocessor; or perform a hysteretic adjustment of the second control effort parameter to slow a transition of the coprocessor to a low power state while a workload associated with the thread group is executing on the processor.

8

8. The asymmetric multiprocessor system of claim 7 wherein the hysteretic adjustment of at least one of the first control effort parameter and the second control effort parameter includes delaying a time between the workload being submitted to the coprocessor and decreasing the first control effort parameter.

9

9. The asymmetric multiprocessor system of claim 7 wherein the hysteretic adjustment includes decreasing a rate at which the first control effort parameter decreases.

10

10. The asymmetric multiprocessor system of claim 7 wherein the processor is selected from the group consisting of a central processing unit, a graphics processing unit, a general purpose graphics processing unit, a neural engine, an image signal processor, and a scaling and rotating engine.

11

11. The asymmetric multiprocessor system of claim 7 wherein the coprocessor is selected from the group consisting of a central processing unit, a graphics processing unit, a general purpose graphics processing unit, a neural engine, an image signal processor, and a scaling and rotating engine.

12

12. The asymmetric multiprocessor system of claim 7 wherein the first and second control effort parameters respectively affect at least one of an allocated subset of cores or execution units and a dynamic voltage and frequency state of the processor or the coprocessor.

Patent Metadata

Filing Date

Unknown

Publication Date

September 14, 2021

Inventors

Aditya Venkataraman
Bryan R. Hinch
John G. Dorsey

Want to explore more patents?

Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.

Citation & reuse

Analysis on this page is generated by Patentable — an AI-powered patent intelligence platform. AI-generated summaries, explanations, and analysis may be reused with attribution and a visible link back to the canonical URL below. Patent abstracts and claims are USPTO public domain.

Cite as: Patentable. “Serialization Floors and Deadline Driven Control for Performance Optimization of Asymmetric Multiprocessor Systems” (11119788). https://patentable.app/patents/11119788

© 2026 Patentable. All rights reserved.

Patentable is a research and drafting-assistant tool, not a law firm, and does not provide legal advice. Documents we generate are drafts for review by a licensed patent attorney.